Osteoporosis and related fractures represent major public health problems that will only increase in importance as the population ages. Several studies have now convincingly demonstrated that aging-related bone loss continues, and may even accelerate, in extreme old age. A better understanding of the factors that may be unique to bone loss in old age may help to refine the types of interventions to preserve bone mass in old age. The present application is a competing continuation proposal from the Framingham osteoporosis study group to examine in detail several risk factors for bone loss in old age using three related cohorts, the original Framingham cohort, The Framingham Offspring cohort and a new Framingham minority cohort. The Original Framingham cohort has been the subject of two previous exams that included measurement of bone density. The Framingham Offspring Cohort was recruited from among the children, and their spouses, of members of the original cohort starting in 1971; 3570 are expected to participate in this osteoporosis study. The Minority cohort is currently being recruited and will consist of 300 subjects (34% black and 66% Hispanic). The proposed studies will extend and expand upon the research group s previous investigations of more traditional risk factors for bone loss by taking advantage of developments in nutritional assessment, as well as findings from cellular and molecular biology of bone, which offer an opportunity to examine risk factors for bone loss that have been less well studied. The primary aim (1) of the proposed studies is to examine the effect of dietary factors on bone health, in particular the effect on bone density and bone loss of consumption choices among common food groups, and the effect on bone loss of specific nutrients that have not been well-studied with regard to bone, including magnesium, potassium, vitamin C, sodium, and vitamin K. The dietary studies will be performed using longitudinal data on bone density from the original Framingham cohort, and using cross-sectional bone density data in the Offspring and Minority Cohorts. Dietary data, in the form of food frequency questionnaires in all three cohorts, and 3 day food records in the Offspring Cohort, have already been collected at previous examinations. For this aim, follow-up for bone loss in the Original Framingham Cohort will be extended from 4-5 years to 8 years by adding an assessment of bone density at a planned future Framingham biennial examination (Exam 24). A special call back visit will be required to obtain bone density in all of the approximately 3,600 members of the Offspring Cohort. Both dietary data and bone density are already being measured in an ongoing examination of the Minority Cohort. There are also several other Aims of the study, as follow: (2) to examine the cross-sectional association of IGF-1 and IGF-BP4 with bone density in a subset of 100 men and 100 women in the Framingham Offspring cohort using blood specimens obtained 5 years prior to the measurement of bone density; (3) to examine the cross-sectional association with bone density of two new measures of weight-bearing physical activity, a validated questionnaire and an automated weight-bearing activity monitor, in subsets of 200 men and 200 women in each of the Original and Offspring Cohorts; physical activity will also be examined in relation to a new measurement of QUS of the heel in this subset from the Original cohort; the physical activity measures will be obtained in a special callback or regular visit in the Offspring Cohort and at Exam 24 in the original cohort; and (4) to compare measures of quantitative ultrasound (QUS) of the calcaneus between members of the Original Cohort who attend Exam 24 and those who are unable to attend the exam to determine if bone loss may be underestimated by studying subjects who attend clinic examinations. QUS will be assessed with a new dry system device in all those in the original cohort who attend Exam 24 as well as those who receive the standard Framingham home visit.